# Tektronix DMM4020

#### Angelo

Sep 10, 2014
30
Dear All,

i am new here, looking for some help since i am trying to repair the instrument in the subject. I honestly say that i get it from a kind of "bid" for few money and trying the repair. So i don't have much information on the origin of the fault.

Anyway,

ISSUE: at power on, an internal relay (on the input measurement part "clicks"), then nothing is displayed out in the front LCD.

What i did until now:
1) measured that all the main PS voltages are there, there are 2 7805, 1 7809, 1 LM399. Output voltages are all there except for 7809 where i measure 11.2.

here a picture

2) i briefly checked the cpu, clock is correct (4 Mhz), cpu is correctly powered 3,3V

3) i moved so the the LCD board, there is a small PS to generate 40V (called high voltages)

On VH i measured initially 40V , but now, after some measurements here and there, i read 24V, (hope i didn't damaged anything additional).
On 9V i measure 11.2
On 5V i measure 5
On EK 3V.

Some questions:
Is there a way that i can try LCD segments separately, so to be sure LCD is working ?

I don't have any service manual also.
Any help on how to proceed is welcome.

Many thanks,

Regards,
Angelo

#### Angelo

Sep 10, 2014
30
Hi all,

above there was an error, there are 2 7805 and one 7905. I measure correctly +5 +5 and -5. The 11.2 V are in a test point where is written 9V. I am considering this voltage ok.

I did some progress:

there is no activity on the LCD driver (clock, data etc), so seems nothing is processed from the CPU, this is why probably the display is always off.

CPU seems is in a locked state, looking the serial data of the near 2 wires serial eprom (AT24C16A), on SCL i don't see any clock.
And strangely, on SDA i see with the scope always the same byte value over time.

#### (*steve*)

##### ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,505
11.2V from a 9V reg is way out of spec.

Please measure the voltage from the ground pin to the input pin, and from the ground pin to the output pin.

OK, if that's not regulated, please determine the source of that voltage.

There is certainly the possibility that the CPU is dead, and that would be hard to source. Is there a part number on it? Check to see that the reset pin isn't constantly asserted or something.

#### Angelo

Sep 10, 2014
30
Hi, thanks,

the 9V being 11.2 is not a problem, they are rectified after a transformer so i don't see any special issue since they are used for the display board only.

Let's come to the hardest part.

The micro is a TI MSP430F449, 16bit RISC, with 60K internal flash and some internal SRAM.
Now, i see that the reset is pulled high correctly, but it is probably hanging in some loop or exception.
- No data seems read for example from the serial eeprom (that should contain some nv config data).

I do still some investigations, i have nothing to lose, then an idea is to dump the code with JTAG, and disassemble it to see if the code is damaged.

Will proceed in some steps tonight.

#### Angelo

Sep 10, 2014
30
Hehe, i also did a damage to the OKI LCD driver chip, since i applied a 30V DC to a led segmet to test it. Now the chip is a bit hot, most probably i damaged the output of the chip connected to the same line. I ordered a replacement chip.

Damaging is part of learning. Once damaged, you should have learned (since you pay).

Last edited:

#### Angelo

Sep 10, 2014
30
Yesterday i had only time to re-check all the power pin of the cpu are correctly powered. I don't see power issues. I dind't see any kind of activity in any pin of the cpu. Likely, cpu is not starting up, looping internally in some init state.

A strange thing, i touched for error + and - of the cpu, and could hear a relay clicking.

So seems this relay could have some protection purpose. So another idea is that the instrument could be in a kind of "protection" state, since i hear the relay clicking just after power on. And that cpu, detecting this from some pin, stops/loops some where.

I am still in a very gray phase. Tonight hopefully i would try to read the cpu program through jtag.

This instrument is the same as Fluke 8808A. Would be nice to find a service manual, a diagram or something, but was able to find anything.

Every help is very appreciated.

#### Angelo

Sep 10, 2014
30

i replaced the OKI display driver, no changes, but almost i repaired damage done by me (LCD high voltage back to 40V).

Still convinced all here is a cpu problem, it "seems" not starting, there is no activity on any pin, i don't see the clock that should drive the LCD shift register (since manual says that at boot all LCD segments must lights up), i don't see data on any pin. _Reset goes high correctly at startup, but with scope in single shot mode i don't see any activity at boot on any pin.

I ordered at OLIMEX for 16E the jtag parallel programmer, so i can use linux "mspdebug" to hopefully read internal flash. Once i have the binary dump, i would be able to program a new cpu.

Some function relays clicks at boot, so could be that cpu perform some activity and then loops somewhere (just assumptions), so i am still investigating.

I also tested all tantalium caps, some bjt transistor here and there, disconnected the fron tanel (lcd + keys), no changes, still i don't see any activity of the cpu.

#### (*steve*)

##### ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,505
Best of luck. You're going to a significant effort and I hope it pays off.

#### Angelo

Sep 10, 2014
30
hi steve,

well, i get it for 100$on ebay, instrument cost is 500 to 1000$ (new), so even if i don't get anyway, once i found another i can make 1 from 2.

But i like challenges firstly for learning and do experience.

Now we are entering on digital issues, that is more comfortable for me, since i can have more issue with analog things.

Only case i can get stuck is if the cpu is damaged and i cannot dump the cpu program, so can't reprogram.

In AVR (atmel forum, general electronic section) a guy told me that the small eeprom stores some config and calibration values, and can be damaged. At the same time, seems there is a procedure to restore default values in the manual.

Yes there are chances i2c eeprom have issues, since strangely, the cpu (master) is not acting any clock on SCL, but i see a gnd level with some small bits over it, so maybe the i2c is not allowing the cpu to drive the signal.

Now i am probably going to try new i2c eeprom + restore defaults.

#### Angelo

Sep 10, 2014
30
Ok, i soldered a brand bew 24C16, nothing changed at signal levels, cpu is not clocking. I did the default "clear" non-volatile procedure, but i guess it hadn't any effect, since display is black and cpu is not running.

Impression is like cpu flash (60K) is empty, that there is no program inside, since nothing appears on any pin at boot.

So, who have the same instrument and can read with a jtag programmer for msp430 the code, and send it to me ? Huauhaa, probably the answer is Tektronix.